>  Yep, sounds right to me. If your digi player was -really-
 > intelligent, it could do a hunt through $1000-$2000 for $18 $d4,
 > and use what it found (-: (nah, too damn hard, you could only
 > assume that the first one found was the init.. (usually is)).
You could make it a user option.. ie: user presses a key, the digi-player
looks through the sid-tune and finds possible locations, then displays them
for the user to accept/discard/modify. The digi-player would have to keep
those values stored away somewhere, so the user can properly test that the
values he accepts work, and if they don't, restore them.

 >  Of course all this discussion is only of use if you are
 > actually going to write a digi-organiser type clone.. (-:
And why not? The only reason I DON'T use digi-organizer is because it isn't
compatible with DMC (haven't tested it with DMC 5 yet though), and it also
did not have any effects.

 >  TM>I'm thinking about it... at the moment though, I'm just trying to find
 >  TM>some time to relax (been a rather stressful few weeks)

 >  Well do it! After you've built a working version, I'll sponge
 > the circuit off you & build my own. (: How you can be stressed
 > when you dont even work is beyond me however.. (-: (depressed,
 > even suicidal I could believe, but stressed? when you have a
 > continuous holiday? (: )
Um, I've lost what we were talking about here... hang about...
oh yes, the cartridge daughterboard! Hmm, I'll see about obtaining a multi-
cartridge type thing (like Russell's) and modifying that.

As for being stressed... you don't want to know. Try managing a house sometime :/

 >  TM>Hey, there's one important factor here... I'm not TBH! ;)
 >  What? Aren't you?? Damn! & I though you were one & the same! (-:
Hmm you could be right, actually, I'd better check! (Has a reality crisis)
Yes, if fact I AM TBH! HA! ... NOT!

 >  Yeah, I figured that.. Although I think the music I am thinking
 > of probably changes waveforms & possibly frequency slightly as
 > well.
Hmm, have you listened to the tune from Coma Light 12? I think this uses
the kind of effect you are talking about...

 >  Volume? Several possible ways, all with disadvantages. On newer
 > 64's & 128's, the signal has to be 'carried' on a 100% pulse
 > width square wave (ie a DC level). You could change the sustain
 > level of the 'carrier' voice to change the apparent volume. Wont
 > work on old 64s tho!
Well, perhaps you can have this as a user definable option...

 >  And you would change which table you loaded from, and the tables
 > would have different 'scales' of volume from 0-15. I think I
 > explained that alright.. (-: ie a 'full volume' table would go
 > from 0-15, but a 'half volume' table would have 16 values from
 > 0-8. So a $0f in the sample would actually only put a $08 in the
 > volume register.
 >  Actually, I've confused myself (: It's not that simple, as
 > values actually range from 0 to 15, 'centred' at about 8. So
 > half volume would actually go from 4-12..
Well, that's actually pretty straightforward. Why didn't I think of that? :)

 >  As I just attempted to explain, the samples are centred at 8,
 > so chopping off the top couple of bits would wreck the sound, ie
 > no you cant do it that way! (I wish you could). Also, LSR'ing
 > wouldnt work (pity)..
What about if you cleared the "outside" two bits, ie: bit 7 and
bit 0? What kind of effect would that create?

 >  As for 8 bits, it would only work on 2 SID's if you could play
 > the lower nibble sid at exactly 1/16th the volume of the high
 > nybble sid, which is of course impossible!
Damn! :) Then again, Sideborders, FLI and Real-time vectors were impossible
on the 64 too ;)

 > The only way to play samples >4 bits on a 64 is to use PWM, which
 > gives you up to 12bits in theory, but the quality sux (:
I remember you were talking about the high pitched noise in the waveform...
couldn't you use the low-pass filter to get rid of it? That would make the
sample a little easier to bear...

 >  btw, Datel's sampler could record & play back at 8 bits, so
 > this commercial product you have found is not original. My home
 > made sampler (hacked amiga design) is actually 8 bit, & I just
 > LSR LSR LSR LSR the sample to get 4 bits out of it. (I'm lazy &
 > couldnt be bothered just rewiring it (-: )
HOW does the datel player replay at 8-bits? This is what's confusing me at
the moment...

 TM>which, do you think you could tell me why all I get from my
 TM>sample is a VERY distorted of what I'm playing through it?

 >  Sounds like you have the signal wires mixed up. I think you
 > told me that you had just connected the high 4 lines from the
 > sampler to the lower 4 bits of the user port. It is possible
 > that a) you got it mixed up & have connected the lower 4 bits
 > from the sampler, or b) you have the wires back-to-front (ie bit
 > 0 connected to bit 3 etc).
I actually connected pins 2-9 (d0-d7) of the parallel port connector on the
sampler to pins C-L (pb0-pb7) of the user port connector. I'll try wiring
it round the other way.

 >  Or even a combination of the above! (: The ADC0802 chip has the
 > pinouts as bit 7 is pin 11, bit 0 is pin 18. The ADC0804 in your
 > amiga sampler has the same pinouts. (to my knowledge). Check to
 > make sure you have it the right way round. Try connecting all 8
 > lines like I have if you are still not sure what is wrong.
 >  btw with nothing connected to the sampler, it should read about
 > 127-128 (or 7-8 in your case) on the user port. There -may- be
 > an adjustment ont he sampler to 'centre' it. Some samplers are
 > fixed though, and the knob only adjusts volume..
No centering knob, but two volume knobs (stereo sampler).
I'll try to re-wire this tonight. I'll get back to you with the results!

 >  Yeah, I can think of an easy way to do that actually, instead
 > of AND #$0F, just self-modify it to AND #$00 (-: Well that's 3
 > effects, keep em coming! (-:
Another might be to try and allow the sample frequency to sync up with
voice three via $d01b/$d01c...

 > btw with the 'looping', I can get the loop start to within a byte,
 > but the end of the loop must be on a page boundary. (the code would
 > just be TOO long if I had to check if it ended every time I
 > incremented the sample address..)
What if you were only doing one sample? Would that shorten the code enough
to do a full loop check?

 >  TM>Perhaps you could write an article for them about sampling?

 >  Haha, I might (: There is a bit of discussion in comp.sys.cbm
 > on sampling at the moment, with some people giving long
 > descriptions of how NMI's work etc. (often incorrectly
 > unfortunately).
Well you should be in there correcting them! :)

 > btw, I got hacking #1 from his mailserver (if you want it).
No, tis cool, I can get it from Russell.

 > btw2, working on a basic pattern/track player for digis at the
 > moment, should have something working by tonight..
Look forward to it.

I've been looking through our version of Macro-assembler, specifically at
the cross-assembling code... and I've written a cross-reciever for it! I'm
finalising it at the moment (trying to work out why/when certain bytes are
sent, ie; for pass 2, and the start-code command), but should have a fully
working version up to you soon (ie: next send )

Tarragon/Menace/Gothic Designs

--- FMail 0.94